disparity between core clock and SMP clock

I’m trying to make some decisions about what to buy, and there is one point i’m having difficulty understanding and finding information about.

Product specifications for the G280 GPU indicates that it the multiprocessor cores run at about 1.3 GHz and the memory frequency is 1.1 GHz. Yet there is also a field called “graphics clock” which runs at 600 MHz. What is this? And how can I calculate peak performans in gigaflops from these numbers?

Thanks alot in advance.

SPs run at the higher (shader) clock value. The core clock value is for other parts of the chip (texture samplers I think are at core clock, rasterization hardware that isn’t used by CUDA, and certainly other stuff I’m not thinking of).

I am merely solving differential equations with values pulled out of memory. Do I need to be concerned about the core clock or will all my work be done on the shaders?

That would only use the shader clock, yes.